淺層地滑發生機率之研究－以大粗坑集水區為例

Predicting the Probability of Shallow Landslide Occurrence-Case Study of Da-Tsu-Ken Watershed

10.29417/JCSWC.201012_41(4).0001

何瑞益(Jui-Yi Ho)；李光敦(Kwan-Tun Lee)

淺層地滑預測 ； 第一階不確性分析理論 ； 水文模式 ； 邊坡穩定分析 ； shallow landslide prediction ； first-order analysis of uncertainty ； hydrological model ； slope instability analysis

中華水土保持學報

41卷4期（2010 / 12 / 01）

285 - 295

繁體中文

本研究藉由第一階不確性分析理論，針對邊坡穩定分析理論所需因子進行不確定性分析，研究中並採用數值高程模式與降雨逕流模式，以計算各項因子之平均值與變異數，進而利用風險分析理論，計算豪雨時期間可能引發淺層地滑之機率。本研究蒐集臺北縣大粗坑集水區之水文及地文資料進行分析，並採用水文模式配合三場颱風之小時雨量資料，以模擬集水區淺層飽和水位之時間與空間變化。研究結果顯示，2000年11月象神颱風於研究次集水區所造成淺層地滑發生機率高達60.31%，其分析結果與災後調查結果大致相符。因此應用本研究之分析方法，應可藉以發佈地滑之即時警訊，以減緩人員傷亡與財產損失。

This study employed the first-order analysis of uncertainty to analyze the model parameters used in the slope-instability analysis. The mean value and the variance of the model parameters were determined using a digital elevation model and a rainfall-runoff model, and then these parameters were substituted into the risk analysis to predict the probability of shallow landslide occurrence. Hydrological records and geomorphologic information from Da-Tsu-Keng watershed in New Taipei City were collected for analysis. Using the hourly rainfall data from three typhoon events, the temporal varying of the mean value and variance of the saturated water table were simulated using a hydrological model. The results show the probability of the shallow landslide occurrence reached 60.31% during Typhoon Xangsang in November 2000. The predicted result was consistent with that obtained from field investigations. The proposed analytical procedure is therefore considered promising for real-time landslide warning to decrease the magnitude of the loss of lives and properties.